Robust aeroservoelastic design with structural and trim parameter uncertainties

A unified robust structural, trim relationship and controller design methodology is presented for a MIMO aeroservoelastic system. It contains two stages in an iterative procedure: (1) robust structure and trim relationship optimization, and (2) robust controller design based on the optimal structure. The objective of the unified robust aeroservoelastic design is to obtain a minimum structure weight, under constraints on stability and performance specifications, and with structural and trim parameter uncertainties. Robustness is reflected in the fact that the weight variance is restricted to a small range and all the constraints are met when the design variables are uncertain. A genetic algorithm and sensitivity data were used in this open loop robust design stage. For the robust controller design, the well-known linear fractional transformation is used to interconnect uncertainty models for the structure and trim parameters with the other nominal aeroservoelastic model. Finally, the μ synthesis is applied to design a robust controller, which provides robust stability and instaneous closed-loop performance, under reasonable stress limitations. A flight vehicle with four control surfaces was tested to validate this methodology. After robust aeroservoelastic design, the aircraft is 3.62% heavier than the structure optimized by nominal method without consideration of uncertainty. However all the constraints are satisfied by the robust optimization, when the design variables are perturbed in a 5% uncertainty range. Hence, the structure is more robust to resist design variable uncertainties. The robust controller provided simultaneous high roll-rate performance and reduced the wing-root stress by 58.3%, when compared with the performance and stress of the openloop aeroelastic system.